1. Field of the Invention
The present invention relates to managing addresses and network bandwidth. More particularly, the present invention relates to managing dynamically allocated addresses and applying in a consistent manner a quality of service scheme which manages network bandwidth associated with the dynamically allocated addresses.
2. The Background
As shown in FIG. 1, the Internet 10 (or another network) may be described as a group of interconnected computing networks (not shown) that are tied together through a backbone 12. The computing networks, in turn, provide access points, such as access points 14, 16 and 18, through which users may connect to the Internet 10 via a station (a computer having a connection to a network) or host, such as hosts 20, 22, 24, and 26. An access point is essentially a location on the Internet 10 that permits access to the Internet 10. An access point may include a modem pool (not shown) maintained by an ISP (Internet Services Provider) which enables its subscribers to obtain Internet access through a host having a dial-up connection. Those of ordinary skill in the art will recognize that other types of access methods may be provided by an ISP such as frame relay, leased lines, ATM (asynchronous transfer mode), ADSL, and the like.
Regardless of the access method used, each device (e.g., a host or router) that receives, sends and/or routes information between or among other devices on Internet 10 is configured to communicate with other devices using a communication protocol that may be understood by the other devices. The current communication protocol used by these devices on the Internet is TCP/IP (transmission control protocol/internet protocol). In addition, each device that can send or receive information (e.g., a host device) must also have a unique address. The type of address used for the Internet, or an equivalent switched network that uses TCP/IP, is commonly referred to as an IP address. A standard TCP/IP address is presently 32 bits in length, providing a total of 232 possible IP addresses. Those of ordinary skill in the art will readily recognize that not all of these possible IP addresses are available due to administrative expediencies, such as reserving blocks of IP addresses for future use.
Sending or receiving information using the TCP/IP protocol requires encapsulating information into packets. Each packet includes a header and a payload. The header contains information related to the handling of the payload by a receiving host or routing device, while the payload contains part or all of the user information. The information in the header includes the sender's and the recipient's addresses and is used to route the packet through the Internet until the packet is received by a host having an IP address that matches the packet's destination address. (When referring to the source address and destination address of a packet, the source address and destination address are commonly referred to as “SA” and “DA”, respectively.) This approach enables users to accurately exchange information with one another through their respective host computers.
By implementing a protocol common to all devices using Internet 10, users may send and receive information with other users on the Internet in a seamless manner regardless of geographic location or the type of host and/or interconnected network used. This ability for users to access or provide information to others regardless of geographic boundaries is one factor fueling the rapid increase in the number of users using Internet 10. This increase has depleted the number of addresses and has increased the amount of packet traffic on Internet 10.
One solution for mitigating the effect of the number of users requiring addresses is to dynamically allocate addresses for users who do not have dedicated connections to Internet 10, such as users who use dial-up access methods to connect to an ISP. Dynamic allocation of IP addresses entails having a pool of IP addresses, such as IP address pool 32, from which an ISP, such as ISP 34 can draw each time a valid subscriber (who does not use a dedicated connection or a connection that does not have a framed IP address, i.e., a static IP address) seeks to access the Internet. Once the subscriber logs on to an ISP and is properly authenticated, the ISP allocates an IP address for use by the user. Upon log-off, the assigned/allocated IP address is then released, rendering that IP address available for subsequent use by another user. In this way, a set of IP addresses can be used to provide access to a number of users that exceed the number of IP address comprising the IP address pool, assuming that at any given time the number of users seeking to log-on and obtain dynamic IP addresses is less than or equal to the number of IP addresses available in the IP address pool.
Attempting to mitigate the effect of high levels of packet traffic on the Internet has resulted in different solutions. One solution, albeit an expensive one, is to increase bandwidth throughout the Internet such as by providing an Internet backbone that has a higher bandwidth, i.e., providing a larger pipe. Another solution is to manage bandwidth used by Internet users according to some defined set of criteria so as to decrease the latency period experienced by a packet while being routed from an access point to an Internet backbone. This approach is commonly known as QoS (Quality of Service). QoS may be thought of as a solution that does not create a bigger pipe through which packets may flow but a solution that picks which packets will go first, i.e., some packets have precedence over others.
Packet precedence may be based on the address used by a user's host machine. For example, a user may have service level agreement with an ISP that requires ISP 34 to give precedence to packets associated with the user. Packets given precedence are selected according to the source address held by each packet. To ensure that precedence is provided consistently, the user is given a static or permanent IP address that when encapsulated in a packet and received by a router, such as router 36, is given precedence over another user that does not have the same service level agreement. This decreases transmission delay experienced by packets originating from the user's host machine when being routed from access point 14 onto the Internet backbone 12.
However, basing QoS or precedence on a static IP address precludes ISPs from using dynamically assigned IP addresses because the same user may have different addresses each time the user logs on. Moreover, if the same user decides to obtain Internet access at an area serviced by a access point (sometimes referred to as a PoP or “Point of Presence”), provided by the same or different ISP, the ISP will not know the user's static IP address. Static IP addresses are not shared between access points, such as access points 14 and 16, that are connected to Internet backbone 12. This precludes the ISP from providing the same QoS to its user simply because the user is a roaming user (i.e. a user that uses an access point that is other than the user's home PoP location) even though the PoP or access point is serviced by the same ISP that services the user's home PoP.
Accordingly there is a need for managing the dynamic allocation of an address to a host device which is selected to obtain network access from any access point within a given communications network. Further, there is a need for maintaining a network bandwidth management scheme that is consistently applied to a user's network bandwidth usage regardless of the access point used by the user.